CSC5160: Combinatorial Optimization and Approximation Algorithms Topic: General Matching Date: 24/01/2008
Lecturer Lap, Chi Lau, Scribe Jennifer +1 more
In this lecture, we discuss matchings in general graph. Following the same structure of bipartite matching, we first introduce the min-max theorem for general matching. Then we present Edmonds’ polynomial-time algorithm to find a maximum-size matching. Finally, we solve the Chinese postman problem using general matching. We have considered matchings in bipartite graph, now we will consider matchings in general graph. Recall that a graph is bipartite if and only if it contains no odd cycles. Given a graph with weighted edges, the goal of the Maximum Matching problem is to find a matching (a set of vertex-disjoint edges) with maximum total weight. In this lecture, we focus on the cardinality version where the goal is to find a matching with the maximum number of edges. 5.1 Min-Max Theorem for General Matching In bipartite matching, we have König’s theorem as the min-max theorem, which…